[all-commits] [llvm/llvm-project] 81e0de: [MLIR][Arith] FastMath extf conversion without NaN...

[Renato Golin via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 81e0de2ea538991a759c921a5d257b271d7d0f7c
      https://github.com/llvm/llvm-project/commit/81e0de2ea538991a759c921a5d257b271d7d0f7c
  Author: Renato Golin <rengolin at systemcall.eu>
  Date:   2026-02-11 (Wed, 11 Feb 2026)

  Changed paths:
    M mlir/lib/Dialect/Arith/Transforms/ExpandOps.cpp
    M mlir/test/Dialect/Arith/expand-ops.mlir

  Log Message:
  -----------
  [MLIR][Arith] FastMath extf conversion without NaN checks (#180926)

This PR allows the expand op converter to consider the NoNaN fastmath
attribute to disable the runtime checks for NaNs in E8M0 types. Default
behaviour is still the same.

The OCP document provides all-ones as NaN for E8M0, but for pre-MX I8
quantization, the checks for NaNs are prohibitively expensive,
especially if the hardware doesn't have native support for that type.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications